强干扰条件下的星体提取方法

摘要

To solve the problem of star extraction on the condition of strong interference, a star extraction method based on edge detection+selection for star pixels was proposed in this paper. An edge detection algorithm to segment the star image is first utilized, and then each pixel is marked. Finally, self-adaptive threshold to select the object pixels of each star is used. Also, the pixel labeling algorithm is improved, and before the pixel of target is labeled, the connectivity of its neighbors is examined so that all of the pixels of the same star only have one mark and the processing efficiency of star image is improved. Experimental results show that the method has better robustness to strange interference, higher speed and accuracy. The improved method can distinguish sensed stars more quickly, and may be more convenient for engineering applications.%为了解决在强干扰条件下星体提取问题，提出了一种基于边缘检测＋星体像素筛选的星体提取方法，首先用一种边缘检测算法分割星图，然后对像素进行标记，最后利用自适应阈值对星体的目标像素进一步筛选。并对像素标记算法进行改进，改进的算法在标记目标像素时，同时分析与它相邻的像素的连通性，使得属于同一星体的所有像素只分配一个标记值，提高了星图处理的效率。试验结果表明，该处理方法对强干扰噪声具有更好的鲁棒性、且运算速度更快，精度也有所提高。提出的边缘检测＋星体像素筛选的星体提取方法有助于在强干扰条件下分离星体目标和背景，改进的像素标记方法，可以更快地区分不同的星体，便于工程应用。